The Research Solutions team has taken several steps to make sure that a responsible and ethical approach is leveraged for the development and deployment of AI in our applications to maximize the benefits to our customers while minimizing the potential risks.
We are committed to ensuring a responsible and ethical approach is leveraged for the development and deployment of AI in Article Galaxy. Our team prioritizes maximizing benefits to our customers while minimizing potential risks.
Here are some common questions and their respective answers you may have about our use of AI.
How is Research Solutions using my user or company data?
Our customers’ privacy is paramount. Therefore, no company information of any kind, nor any entity identifying data, is used in any of the interactions with AI.
How is Research Solutions using our search query data?
No user search query is passed to any AI engine. To improve the accuracy of the search, the only data used is generic article citation data, such as document title, publication name, abstract, DOI, author(s), or other related citation metadata.
Are my company’s users able to interact with the AI engine directly?
No, the AI integration within Article Galaxy is restricted to Article Summarization and the user has no ability to input or modify any of the prompts. These pre-determined prompts were purposefully created and tested to ensure the generated response would be limited to only the prompt query.
Users do not have the capability to engage in direct chat with the AI or generate their own prompts.
Is Research Solutions using any of my company’s content (e.g., articles) to help train any AI model?
No, our AI integration is focused on data summarization only; we do not pass any proprietary data from Research Solutions or our customers to train any AI model.
How is Research Solutions ensuring the accuracy and authenticity of provided data?
Our AI interactions are structured using a method called Prompt Engineering. This method builds a clear set of guardrails for the AI engine to make sure that the data it is evaluating is only within the context of the query and effectively mitigate data hallucinations. Given that our use case is Article Summarization, we are currently employing several parameters within our prompts to do this. These prompt parameters are made up of publicly available information and in no way expose any customer data. All the following parameters are being leveraged for Article Summarization queries:
| Title | Publication Name | Publication Year | ISSN |
| Abstract | Authors | DOI | PubMedID |
These parameters help the AI hone in on the correct document data to summarize in order to provide the most accurate result.
From startups to our enterprise customers, we remain dedicated to customer data security and privacy, implementing the best in safety measures and precautions to deliver confidence and peace of mind.
Feel free to contact customersupport@reprintsdesk.com with any questions or request a free eBook on generative AI here - https://www.researchsolutions.com/ebook-chatgpt